Personalized Nutrition and the Microbiome

The intersection of personalized nutrition and the human microbiome represents an exciting new science. Within this emerging and complex arena, the human microbiome and the impact of dietary components has been the target of approximately 4,000 publications over the past four decades. The field of paleomicrobiology recently reported—based on findings from frozen and mummified human remains as well as desiccated feces and dental calculus—that the ancient microbiome provides an interesting portrait of early dietary behaviors and perhaps intercontinental migratory patterns (Warinner et al. 2015). Some of the microbiota findings suggest humans may experience diverse seasonal variations and other perturbations, such as antibiotic treatments and geographical settings, which may also impact modern and mobile societies (Smits et al. 2017).

Animal models, short-term intervention studies, and epidemiological evidence suggest modulation of the gut microbiome through dietary interventions may impact a variety of noncommunicable diseases, such as obesity, diabetes, inflammatory conditions, cardiovascular health, and some forms of cancer (Holmes et al. 2011). A more recent review noted variations in bacterial genera, such as Bifidobacteria, Lactobacillus, Bacteroides, and Clostridium, are associated with numerous physiological changes that may impact clinical outcomes and general health (Singh et al. 2017, Li et al. 2017).

Additional evidence indicates that diet and genomes interact in complex fashion. While the emerging discipline of nutrigenetics is shedding significant illumination on how an individual’s genetic makeup predisposes for dietary susceptibility, nutrigenomics is clarifying the ways in which nutrition influences the expression of the genome and its impact on health outcomes. For example, the management of conditions such as those associated with metabolic syndrome may represent opportunities at the personalized nutrition and human gut microbiome intersection (de Toro-Martin et al. 2017). The shaping of one’s ability to cope with some forms of cancer (Seidel et al. 2017), the metabolism of bioactives (Koppel et al. 2017), the management of gut health (Panduro et al. 2017), and even maturation of an infant’s immune system (MacPherson et al. 2017) are also at this complex interface. Those in the Gordon Laboratory at Washington University in St. Louis have been studying the influence of specific bacteria on energy metabolism and the predisposition to obesity for more than a decade (Turnbaugh et al. 2006). These data from genetically obese mice (leptin deficient model) suggest differences in distal microbiota, namely Bacteroidetes and Firmicutes, may contribute to the onset or pathogenesis of obesity and our understanding of the complexities of microbiota and health.

One’s genetic makeup and genome arguably encompass far more than our set of chromosomal DNA (i.e., inheritable traits). Arguably, it may be said to include noncoding DNA, the genetic material in mitochondria, the DNA or RNA in viruses, and bacteria resident in the microbiome. The specific roles of all this extra-chromosomal genetic material in human development and physiology in health and disease has become an integral theme in the excitement over personalized nutrition.

Those in this research intersection typically seek to elucidate the gene, protein, and metabolic profiles of individuals in varying states of health and illness. It is hoped that correction of deviations in homeostasis can be affected by nutritional means, thereby maintaining health and reducing the risk of disease (Kussmann et al. 2006). At the center of this omics approach is the discovery of appropriate and robust biomarkers that clearly indicate or predict one’s health outcomes and the unearthing of bioactive or beneficial food components that align with one’s genetic profile and phenotype.

One recent publication proclaimed that ostensibly normal/healthy adult subjects can “choose the foods that are right for you by measuring your own personal postprandial glucose response” (Korem et al. 2017). In this case, following a short-term, crossover study (7 days) among 10 subjects who consumed refined white and whole-grain sourdough bread (50 g of carbohydrate from each) did not uncover any significant changes in the gut microbiome, except for possibly two taxa. None of the clinical parameters studies were affected by either dietary intervention. However, there appeared to be person-specific postprandial glycemic responses. Thus, the authors concluded that this interpersonal response variability and subtle changes in microbiota could be used to predict glycemic responses.

Upon critical evaluation of any study, especially short-term trials, one needs to carefully examine methods and statistics. In this case, the macronutrient difference between the two types of bread would not be expected to produce pathologic or even clinically significant differences in levels of serum glucose, especially after a study duration of one week, especially among healthy adult subjects. Additionally, there was no documentation of dietary patterns either preceding or during the study. Within the realm of clinical chemistry, the method of glucose determination for glycemic response to a single dietary modification may not be validated or represent significant variations (Rebel et al. 2012, Ginsberg 2009). Of course, the small sample size and nonparametric statistics present additional threats to the study validity.

Despite interesting preliminary data with intriguing speculation, the assertion that the subtle, yet insignificant changes in the microbiome modify the character and magnitude of a glycemic response in a clinically significant fashion, particularly in healthy adult subjects, is not at all convincingly supported by the evidence. Clearly, it is premature to even speculate that the apparent genetic signature of an individual’s gut microbiota can be used to “optimize food choices worldwide.” Even if a person-specific microbial genome can be utilized to predict that individual’s glycemic response, it is less than apparent how that information could be utilized to guide some optimal adjustment in diet in a healthy adult involving a single dietary component much less for the complex matrix of food that represents the diet in a free-living adult.

Despite the hyperbole advanced in the bread study, the development of tools that support—and may ultimately help guide—personalized nutrition, and more broadly, personalized medicine, is clearly an important pathway for investigation. The research is promising. And the standard of practice within the medical field is common, especially with respect to chemotherapy. In addition, epigenetics is demonstrating its possible implications in nutrition, including the imprinting effect that nutrition can exert on a genome, especially during critical periods of growth and development (Junien 2006). Equally noteworthy is DNA methylation, which appears to provide a format for long-term dietary reprogramming of the genome (Waterland and Jirtle 2003). This topic was addressed in two separate articles in this column a decade ago (Clemens and Cheng 2007a, Clemens and Cheng 2007b).

As scientists and consumers, we must acknowledge and respect the complexity of nutritional systems biology, and we must not allow our idealism to exaggerate the impact of research findings that may either support or contradict our understanding of unique problems, patterns, history, preferences, and expectations in the applications of personalized nutrition and microbiome modulation.

Roger Clemens, DrPH, CFS,
Contributing Editor
Adjunct Professor, Univ. of Southern California’s School of Pharmacy, Los Angeles, Calif.
[email protected]